training_config.json

{
  "record_count": 0,
  "restarts_per_sample": 2,
  "episodes_dir_initial": "persistent-data/state-service/episodes",
  "episodes_dir_final": "persistent-data/state-service/shallow-zero-style-episodes",
  "checkpoint_dir": "logs/_training_artifacts",
  "pv_dir": "persistent-data/trainer/models",
  "device": "cpu",
  "backbone": "lg_transformer",
  "bc_epochs": 5,
  "ppo_epochs": 40,
  "post_recording_ppo_epochs": null,
  "h_max": 300,
  "w_max": 160,
  "start_port": 3003,
  "max_assistant_actions": 1,
  "human_follow_up_actions": 1,
  "record_timeout_seconds": 600,
  "assistant_noise_prob": 0.2,
  "assistant_noise_top_k": 3,
  "record_poll_interval": 5.0,
  "state_service_url": "http://localhost:8000",
  "tester_wait_seconds": 600.0,
  "tester_poll_interval": 5.0,
  "train_extra_args": []
}